A novel technology for crop monitoring. How does it work? And what are the benefits?
Hyperspectral remote sensing is the latest advancement in crop monitoring and precision agriculture. It allows farmers to ‘see’ in their crops and gives them the knowledge what the conditions of their crops are, what biological processes are occurring and what actions are need to be made before going into their fields. But before going in-depth on hyperspectral remote sensing. What is remote sensing in the first place?
What is Remote sensing?
Remote sensing can be best described as: “The science and art of obtaining information about an object, area, or phenomenon through the analysis of data acquired by a device that is not in contact with it.”
Its applications dates back to world war 1, when it was crucial to know what happened behind enemy lines, how cities and infrastructure looked like and where strategic places were localized. This information was acquired by attaching cameras to homing pigeons, which were already used to convey written messages. From these images, soldiers could get valuable information about the ground situation and to know the best timing to succeed in their missions.
Nowadays, the ‘pigeon-technology’ has improved immensely. Our vehicles of sensing technologies are: satellites, airplanes, drones, robotics, hand-held devices and sensor technologies. Each having its own benefits. A satellite or airplane can cover a large area quickly, but with little intervals, low resolution and little detailed information. Information from satellites or airplanes remains therefore sample based and is little actionable. A robot or sensor however, can gather specific and very relevant information about an object with many intervals, but it is mostly stationary or covers small areas.
Also, we do not have to solely rely on 2D images anymore. New technologies are still being developed to gather more information of an object remotely. One of these major advancements is the commercial use of spectroscopy. Spectroscopy allows the user to measure the energy of a specific wavelength (color) within the electromagnetic spectrum that is emitted, absorbed, or scattered by an object at one single point. These spectral signals can give information about a property or feature of an object.
What is Hyperspectral imagery?
Normal cameras can see 3 bands of colors within the visible spectrum: red, green and blue. Instead of just these three colors, hyperspectral imagery uses specially engineered sensors that measure the entire spectrum of light in the visible, but moreover also in the non to human visible spectrum for each pixel in an image. The entire spectrum is then divided into 100+ color bands.
Each band is analyzed and used to make a multi-spectral colored dataset in the shape of a cube that contains three dimensions of information about a subject: 2 dimensions of spatial information and a 3rd dimension of spectral information. By analyzing all colors of the spectrum in each pixel, we can obtain the unique spectral ‘fingerprint’ of an object. These fingerprints deliver very detailed information about the state and constitution of the imaged object.
There a multiple applications of hyperspectral imagery in many different industries such as: Image guided surgery, mineralogy, optical sorting, food quality inspection and forest monitoring. Still, agriculture and precision agriculture is the largest sector for hyperspectral imagery where most of the developments are being made.
What are the benefits?
Especially in agriculture, most of the inspection relies on human inspection. As mentioned, the human eye can only see a small portion of visible spectrum. There is thus a clear limit to what we as humans can view in the world around us. For us humans two leaves can look identically green. But when using hyperspectral imagery we obtain more information in the non-visible spectrum which allows us to still distinguish between these two leaves on their properties.
Moreover, things as plant stress as nutrient deficiency and diseases become visible to the human eye when the damage is already done. Using hyperspectral imagery, the spectral specific fingerprints of plant stress become visible in the most early stages. This allows farmers to be more efficient with resources, achieve higher yield and prevent crop losses and ultimately reduce their environmental footprint.
At Polariks, we are specialized in high resolution hyperspectral remote sensing and continuous crop monitoring. Interested to find out what potential hyperspectral imagery has for your organization? Please contact us at firstname.lastname@example.org.